Power amp under development

Hi chico67,
I'm sorry to say, you absolutely do need an oscilloscope to figure out what is wrong here. Otherwise you're just guessing in the dark.

Do you know anyone who has access to an oscilloscope? Can you buy a used one that works, and a 10X probe?

-Chris
 
It's been a while I did not build a new amplifier.
Using SMD part and IRFP240 since i have few of them lying around

Partial finish, still waiting heatsink for the other channel.
Feed it using switching power supply @ +/-46V since my target about 200W @ 4 Ohm. It working flawless.

Thank's Quasi for your circuit :worship:
 

Attachments

  • DSC_1829A.jpg
    DSC_1829A.jpg
    329.7 KB · Views: 671
  • DSC_1826A.jpg
    DSC_1826A.jpg
    424.5 KB · Views: 670
  • DSC_1834A.jpg
    DSC_1834A.jpg
    332.2 KB · Views: 657
  • DS1Z_QucikPrint41.png
    DS1Z_QucikPrint41.png
    37.6 KB · Views: 636
Hi crt,
A bad idea. They are not designed for an AC signal and will try to maintain whatever current you set them up for. So not a reasonable load for an amplifier. These devices are designed for testing power supplies, not bipolar AC generators (your amplifier).

You can buy 8 ohm 100W resistors very cheaply these days. I just bought four for $15 approx. So for an 8 ohm load, use four of these per channel and you will have an 8 R load rated for 400 W. You must mount these on a heat sink of course!

I use the "proper" 8 R, non-inductive, 250 W dummy loads dictated by various audio distributors for doing amplifier repairs. Mine are probably about 30 years old now. They were expensive, each stereo set cost me $150 including the heat sinks and I bought 3 sets back then. So what I just suggested for you is amazingly cheap.

-Chris
 
I am looking for a topology to use two of these FETs in an output stage. The amp in this thread looks like a possible topology.

Here is the power FET
http://ixapps.ixys.com/DataSheet/DS100728A(IXTN660N04T4).pdf

The intent is to have an amplifier that can supply unlimited current at low voltage.

I have zero amp design skills and zero pcb layout skills. I can characterize the devices and build amps and buy boards and parts.

Is anyone from this thread interested in participating?
 
Hi crt,
Sorry, missed your reply. I found this link at Digikey. Good for information only.
HS250 8R F Ohmite | Resistors | DigiKey
At Mouser:
NH2508R000FJ01 Vishay / Dale | Mouser Canada
silly price!
HS250 8R F ARCOL / Ohmite | Mouser Canada
Much better. They haven't gone up in over 30 years, so these are a real bargain!

If anyone goes this route, the heat sink must be milled flat to mount the resistor. After blowing one up, my load assembly is getting temperature readouts installed. I have nearly burned them out on a few occasions. You know that when you get that funny hot electronics smell from the load heat sink.

Aside from the lower prices on these than I expected, I feel they are well worth the money if you ever have to make a power measurement that has to be accepted by any manufacturer, distributor or anyone else in the audio field. They are a standard.

-Chris
-Chris
 
if you want more then 125 watts, Class D becomes the only way to go.
stability compensation, Distortion, bias current and output waste heat become troublesome.
I getting 75w RMS out of my amp with 40 volt rails at .0002% distortion

on the schematic R20 R21 R22 should all be equal.
More importatly T9 and T10 should be connected to the rails after the fuse. Currently the driver will destroy the mosfets by popping the gate voltage if either 5A fuse blows.
 
Last edited:
Hi stocktrader200,
if you want more then 125 watts, Class D becomes the only way to go.
stability compensation, Distortion, bias current and output waste heat become troublesome.
I getting 75w RMS out of my amp with 40 volt rails at .0002% distortion
A personal opinion. The technology isn't what the thread is about for one. Also, class "D" hasn't reached a level good enough for listening at home, although it appears that you have settled quite happily for it.

I strongly disagree with you on your point of view. I would say that class "D" is one way to go, you are claiming that it's the only way that makes sense. Your statement being too exclusive really.
I getting 75w RMS out of my amp with 40 volt rails at .0002% distortion
Probably not true, measured performance to be honest with you. Plus, you need a special filter in order to get any reading due to the switching transients that good instrumentation will pick up.

-Chris